So I thought the issue had been solved, but perhaps not. Bruce, we (i.e. David) think the problem with Rouhi’s data is that FreeSurfer did not detect that the original images were tipped forward 90 degrees and as a result the surface registration failed intermittently (i.e. in some subjects it okay, but in others it is not). These data were created with a fairly old version of FreeSurfer (Rouhi can give the version number), and then a significant amount of manual effort was put into defining some retinotopy on the data. Therefore it isn’t really a feasible option to just rerun the data with a newer version of FreeSurfer, but we want to make use of accurate FreeSurfer surface registration to make group average retinotopy results. So the question is how do we go about making sure the spheres are initialized to have the correct rigid orientation before the registration proceeds.

Thanks,

Matt.

From: freesurfer-bounces@nmr.mgh.harvard.edu [mailto:freesurfer-bounces@nmr.mgh.harvard.edu] On Behalf Of Rouhollah Abdollahi
Sent: Friday, August 12, 2011 12:38 PM
To: Bruce Fischl
Cc: freesurfer@nmr.mgh.harvard.edu
Subject: Re: [Freesurfer] Label2label on fsaverage

I tried to rotate the surface by tksurfer and save it as new file . But when I open the new surface like white it was not rotated.
Is there any special thing that I have to do?

Best
Rouhi

On Fri, Aug 12, 2011 at 2:24 PM, Rouhollah Abdollahi <roohy466@gmail.com> wrote:

Dear Bruce
If I want to rotate sucli pattern too if it is the same?

On Fri, Aug 12, 2011 at 1:10 PM, Bruce Fischl <fischl@nmr.mgh.harvard.edu> wrote:

I guess you could rotate the surfaces, save them, then resample the labels so that they get the new surface coords and sample them into the volume, then onto the newly created surfaces. Not sure, we haven't ever had to do anything like this.

On Fri, 12 Aug 2011, Rouhollah Abdollahi wrote:

Dear Bruce
I think the easiest way is to create all surfaces again but with corrected
and well aligned brain. The next step for me to create new labels from old
labels based on white surface (90 degree rotated) to new white surface. But
how? Is there any way to do that? I mean like label2label or surf2surf if
two surfaces are the same but one of them only rotated?!

Bests
Rouhi

On Fri, Aug 12, 2011 at 12:00 PM, Bruce Fischl <fischl@nmr.mgh.harvard.edu>
wrote:
I don't see how anything would work if it was off by 90deg, but
I guess you should try it

On Fri, 12 Aug 2011, Rouhollah Abdollahi wrote:

Because I am using a result from a very old experiment
that done on wrong direction and the orig.mgz s are not
oriented according to
freesurfer template. I cannot creat white because all
results that I have (label files) are on white surface.
The only way that I think I need to
realign all surfaces and volume and apply that rotation on
my label files.


On Fri, Aug 12, 2011 at 11:42 AM, Bruce Fischl
<fischl@nmr.mgh.harvard.edu> wrote:
   not easily. Why would you need to do such a thing?

On Fri, 12 Aug 2011, Rouhollah Abdollahi wrote:

   Dear Bruce

   I have a data set which I can not create the white
surfaces because all of my results are according to white
surface. I would
   like to know is
   there any way to rotate all surfaces and volumes and
labels ... (everything) after finishing recon-all process.
like 90 degree
   around X axis?
   and apply this rotation to all data set?

   Best
   Rouhi

   On Fri, Aug 12, 2011 at 9:17 AM, Bruce Fischl
<fischl@nmr.mgh.harvard.edu> wrote:
      Hi Rouhi

      I don't really understand. Are you saying you
transformed the initial image before giving it to
recon-all? There is a
   switch to
      mris_register to initialize with the tal xform,
but we never use it as in pretty extensive tests it didn't
seem to help.
      mris_register has a very large initial capture
range due to it's global search, so I would be surprised
if the initial
   orientation
      was the problem.

      cheers
      Bruce

      On Thu, 11 Aug 2011, Rouhollah Abdollahi wrote:

      Dear Bruce
      I think I found the problem. The problem came
from the original oriantation of T1 image and I think
mris_register could
   not
      find it itself during
      the recon-all. I added to have initial alignment
to inflated and the problem solved.( mris_register -curve
-inflated
      ?h.sphere <template
      target> ?h.sphere.reg ) I think it is very
important to add it to recon-all because if the user
doesn't know the real
      oriantation it might make
      such a problem.

      Thanks for your quick response.

      On Thu, Aug 11, 2011 at 10:45 AM, Rouhollah
Abdollahi <roohy466@gmail.com> wrote:
         Dear Bruce

         I am working on Freesurfer 2 caret meshes.
I am using   FreeSurfer2CaretConvertAndRegisterClean.sh
from Van Essen lab
   to
      create caret
         surfaces. when I look at my subjects'
surfaces I can see in flat map in caret some of subjects
sulci patterns are
   really
      distorted
         and I think it is related to
sphere registration to fsaverage sphere. Beacsue the code
is using sphere.reg to create
   all
      surfaces.
         for the subjects which I have less
distortion on sulci pattern I have good data (3 subjects)
but for the (9
   subjects)
      rest it seems
         they have problem to register to fsaverage.
Is there any way that I can be sure that
sphere.reg properly created?

      Thank you very much

      Rouhi

      On Thu, Aug 11, 2011 at 4:37 AM, Bruce Fischl
<fischl@nmr.mgh.harvard.edu> wrote:
         Try using fsaverage as the trgsubject
         Cheers
         Bruce

         On Aug 10, 2011, at 8:35 PM, Rouhollah
Abdollahi <roohy466@gmail.com> wrote:

         > .V1_reg.label --regmethod surface
--srcsurfreg sphere --trgsurfreg sphere.reg --hemi rh and
when I tried to load
   the
      output
         label file on fsaverage surface like
inflated one it shows a lot of point looks like it is not
deformed properly. Is
   it
      related
         to registration to fsaverage or something
esle? The question is how I can register the subjects
surfaces to fsaverage
      surfaces?
         -- With Best Regards Rouhi
--0015175120c69a3c6804aa2ffb81 Content-Type: text/html;
charset=ISO-8859-1
         Content-Transfer-Encoding: quoted-printable
Dear freesurfer experts
         >
         > I've tried to export my label file from
individual space to fsaverage space by using this command:
         >
         > mri_label2label --srclabel rh.V1.label
--srcsubject SUB1 --trgsubject SUB1 --trglabel rh.V1_reg.
      > like it is not deformed properly.
      > Is it related to registration to fsaverage or
something esle?
      > The question is how I can register the
subjects surfaces to fsaverage

      The information in this e-mail is intended only
for the person to whom it is
      addressed. If you believe this e-mail was sent
to you in error and the e-mail
      contains patient information, please contact the
Partners Compliance HelpLine at
      http://www.partners.org/complianceline . If the
e-mail was sent to you in error
      but does not contain patient information, please
contact the sender and properly
      dispose of the e-mail.

      --
      With Best Regards
            Rouhollah Abdollahi

      --
      With Best Regards
            Rouhollah Abdollahi

   --
   With Best Regards
         Rouhollah Abdollahi

--
With Best Regards
      Rouhollah Abdollahi

--
With Best Regards
   Rouhollah Abdollahi

With Best Regards
Rouhollah Abdollahi

--
With Best Regards
Rouhollah Abdollahi